hbase-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Xinan Wu <wuxi...@gmail.com>
Subject Re: Scanner Performance Question / Observation
Date Mon, 03 Aug 2009 20:58:29 GMT
6 sec isn't crazy with 0.19. If you really want to research it, have a
look at where the time is spent, creating scanner or actually doing
the scanning. I think it's the former. That being said, upgrading to
0.20 is a much quicker solution. Scanner has been optimized in the new
version.

On Mon, Aug 3, 2009 at 12:01 AM, Kyle Oba<kyleoba@gmail.com> wrote:
> Hey folks,
>
> I have a question (possibly naive) about a scans performance.
>
> My scan is taking about 6 seconds to do the following:
>
> 470 rows extracted
> total size of all rows together is about 1.4 megs
>
> I'm using a InclusiveStopRow filter to limit rows which are being
> extracted by the scanner.  The problem for me is that I will soon be
> scanning 30x this amount of data soon, and 6x30 seconds will be quite
> a hit.
>
> We're already planning to jump to HBase v0.20 soon.  But, wanted to
> see if folks thought the 6 second timing on this scan was reasonable,
> or horribly off.
>
> Thanks.
>
> Kyle
>

Mime
View raw message